Discover, rate, and discuss biology preprints.

SurvMarker identifies survival-associated molecular features using PCA-based weighted scoring and provides an R package with documentation and examples.

This study compares protein language models (pLMs) with simple baselines in combinatorial protein design, finding no consistent advantage for pLMs in predicting or generating superior variants.

TopoFuseNet improves drug synergy prediction by integrating hierarchical molecular structures and multi-scale topological features, outperforming existing methods on benchmark datasets.

The paper analyzes NMR chemical shift perturbations in protein complexes to show that structural plasticity is widespread, with 74% of significant CSPs occurring non-locally to the binding site.

The paper introduces StrAcTable, an automatically generated dataset of protein-ligand complexes with detailed annotations, enabling efficient training and validation of structure-based molecular design methods.

The paper introduces a unified SDE framework for efficient stochastic transcription trace generation, showing high accuracy with reduced computational cost.

metaJAM provides a scalable, user-friendly Nextflow pipeline for metagenomic ancient DNA analysis, improving taxonomic accuracy and workflow flexibility.

STAT is a multi-agent framework that improves spatial transcriptomics analysis by enabling interactive, transparent, and efficient multi-task analysis across different data resolutions.

PHYFUM reconstructs the evolutionary history of normal and pre-malignant tissues using methylation data, revealing early gland divergence in the gut and later divergence in the endometrium.

geneSync harmonizes gene symbols across RNA-seq datasets to improve integration quality by resolving nomenclature inconsistencies.

The paper introduces MANGO, a network-based method to detect gene set clustering by conditioning on network topology and gene degree, showing improved performance over existing approaches.

ORBIT aligns cross-species coexpression network embeddings using orthogonal Procrustes rotation, achieving higher cross-species retrieval quality than autoencoders.

This paper evaluates the performance of linear combination tests for gene set enrichment using different covariance estimators in unbalanced studies, finding ridge estimation more reliable for unbalanced designs.

SCOTCH improves isoform-level gene expression characterization using long-read single-cell RNA sequencing by refining sub-exon boundaries and reducing false positives.

The paper develops BERT-T6, a model using transfer learning on ProtBert embeddings to predict T6SEs with high accuracy and performance metrics.

The paper introduces immuneKG, a knowledge graph framework for immune-mediated diseases, achieving high performance in target discovery using a graph neural network and novel features.

KG-Bench benchmarks graph neural networks for drug repurposing using the Open Targets dataset, finding TransformerConv achieves the highest performance.

This paper proposes a peptide-based TB vaccine using computational vaccinology to identify conserved epitopes across TB strains, aiming for broad-spectrum protection and universal coverage.

MUSE integrates multi-omics data across species to align cellular states while preserving species-specific features and regulatory module conservation.

The paper introduces a method to quantify biological shape heterogeneity using a stratified Wasserstein kernel, demonstrating its application to 2D and 3D biological datasets.